ND Dept. of Health, Division of Air Quality Comments (prepared by Adam M. Miller, Environmental Scientist) on: 

"Update on Proposed Rules and Draft Guidelines for the Oil and Natural Gas Industry" 1:30-3:00pm (Central Time) conference call on 11/30/2015.

There was some talk about the appropriate frequency of optical gas imaging (OGI) inspections by producers.  It does not appear an appropriate OGI inspection frequency has been determined, and that producers could "get credit" for not finding leaks by reducing the frequency of required OGI inspections at facilities that were not found to have leaks.  This all sounds good in theory; however, there are perhaps some factors not yet considered.

From experience using the FLIR GF 320 camera at oil and gas production facilities:
The ability to detect leaks with an OGI camera depends on several factors, not the least of which is the skill of the camera operator using the camera.  Most cameras will have an `automatic' setting but this automatic setting is not good at finding what may be considered "small leaks".  The `manual' setting allows an operator to tune the camera to the environmental conditions to effectively identify volatile organic compound (VOC) leaks, but requires a degree of skill.
Another factor that contributes to whether or not an operator of an OGI camera is able to find leaks is the weather conditions, and location of the sun at the time of the inspection.  The camera sensors would be ruined by looking directly at the sun with the camera.  This sometimes limits where a camera operator can be standing in relation to a potential emission source (tanks, piping, elbows, valves, etc) and in some cases will prevent the camera operator from detecting certain minor leaks.  The inability to detect leaks is also amplified by cold (32 degrees F and below) and windy (20+ mph wind).  The cameras do not work well in adverse weather conditions (cold, wind, rain, or snow).  Adverse weather conditions are typical of North Dakota winters.  In practice, some days are better than others for using the camera to detect leaks.    
The absence of fluid in storage tanks may also contribute to an inability to detect leaks.  It is difficult or nearly impossible to detect leaks using OGI shortly after trucks drain production fluid out of the storage tanks.  The facility may have leaking equipment but because of the timing of inspection (storage tanks without fluid) no leak will be detected.  
Complicating our ability using OGI to determine whether a facility has VOC leaks or will leak in the near future is the useful life of some of the equipment (seals and gaskets).  It is possible that a facility with no detectable leaks one month may be a major source of fugitive emissions the following month due to equipment failures.  Degradation of thief hatch gaskets seems to occur regularly.  It is unclear how often these types of seals should be changed out, but it would be reasonable to require every producer to implement a preventative maintenance program (based on an average useful life) for seals to give some assurance that these seals are more-likely-than-not in good condition and capable of functioning properly.
OGI is a great tool, but it has limitations.  The danger of relying on only OGI inspections as "The Standard" for preventing leaks is that it may cause producers to wait until a leak is detected before action is taken.  Operating a "Find-and-Fix Program" without a preventative maintenance component has the potential to increase fugitive emissions overall.  In such as case, if a leak develops shortly after an OGI inspection, then it is possible nobody corrects the malfunction until it is revealed during the next OGI inspection (which may occur on an annual, semi-annual, or quarterly basis).  

Additionally, allowing for a reduction in the frequency of inspections for companies that do not find leaks at their facilities incentivizes companies either to falsify reports, or conduct OGI inspections poorly so as not to detect leaks which would be opposite of the goals.  Maintaining a standard without allowing for the relaxation of the standard is an important part of operating an adequate leak detection program.  Clearly defining the inspection frequency for all facilities, and not allowing a change in frequency for good behavior, would eliminate confusion that may develop relative to timing of OGI inspections.  Any incentives must be in line with EPA's goals and not counter to them.   

       

